data_final_v2.xls
Re: FWD: data question
November 11, 2016 // WSN, Monterey, California
Doing reproducible science
- Better science in less time using data science tools
- Lowndes et al., in prep
- reproducibility is fundamental, but rarely tested
- tools have changed how we do science
- reproducibility, collaboration, communication
Data science
Data science:
a discipline "to turn raw data into understanding, insight, and knowledge" (Grolemund & Wickham 2016)
Data science
Data science:
a discipline "to turn raw data into understanding, insight, and knowledge" (Grolemund & Wickham 2016)
Conflated questions
Science:
- what is X's habitat preferences?
- when is Y most abundant?
- how wide is their distribution?
Data science:
- how do I import my data?
- how do I subset the years I want?
- how do I visualize this?
Data science tools
Ocean Health Index
a method to score benefits that oceans provide to people
Ocean Health Index
a method to score benefits that oceans provide to people
Reproducible workflow
Reproducible workflow
Reproducible and open workflow
this talk: jules32.github.io/opensci-talk
written in RStudio's RMarkdown, versioned with Git, displayed/shared with GitHub
You can do better science in less time:
Get to your science questions sooner
1. Learn to code
   - in R
   - with RStudio
2. Use version control
   - git
   - with GitHub
   - through RStudio
You can do better science in less time:
Get to your science questions sooner
1. Learn to code
   - in R
   - with RStudio
2. Use version control
   - git
   - with GitHub
   - through RStudio
3. Learn in an intentional way
in a panicfeel empoweredfor a single purposethink aheadin isolationwith a community
Great resources
Free online books, trainings, and webinars:
Recent academic publications:
- "How open science helps researchers succeed" (McKiernan et al. 2016)
- "Best Practices for Scientific Computing" (Wilson et al. 2014)
- "Good Enough Practices for Scientific Computing" (Wilson et al. 2016)
- "Practical Computing for Biologists" (Haddock & Dunn)
Thank you
to the OHI team, colleagues, #rstats community
email: lowndes @nceas.ucsb.edu
twitter: @juliesquid
website: https://jules32.github.io
talk url: https://jules32.github.io/opensci-talk
NCEAS is hiring: nceas.ucsb.edu/positionsopen
Ocean Health Index
a method to categorize benefits that oceans provide to people
scores are modeled using existing data; data intensive
method can be tailored to different geographies
can help inform policy decisions, especially when repeated
Conflating questions
We could reproduce our science questions:
- how does X change over time?
- how does X change with Y?
- how to model Z?
Data science:
- how to subset just the years we want?
- how to transpose columns to rows?
- how to do this reproducibly?
OHI Global Assessments
We expected to easily reproduce our previous work. We had planned ahead:
- coded models
- 130 pages of published supplemental material
- internal documents and notes
OHI Global Assessments
We expected to easily reproduce our previous work. We had planned ahead:
- coded models
- 130 pages of published supplemental material
- internal documents and notes
…But we struggled to reproduce our work.